Date: Martedi' 25 giugno 2002
Time: 17.00
Place: Sala conferenze

Speaker: Prof. Brian Barsky, OPTICAL Project, Computer Science Division and School of Optometry, University of California, Berkeley
Title: Vision-Realistic Rendering

Abstract. We introduce a new concept, called "vision-realistic rendering", a three-dimensional rendering algorithm that simulates the vision of a subject whose optical system is measured using wavefront aberrometry.

The algorithm uses an input depth map to stratify an initial image into disjoint depth plane images, extends these depth plane images, convolves them with a special object-space blur filter, and composites them to form a final vision-realistic rendered image.

Vision-realistic rendering has many applications in optometry and ophthalmology. Such images could be shown to a patient's eye doctor to convey the specific visual anomalies of the patient. Also, images could be generated using the optics of various ocular conditions, which would be valuable in educating doctors and patients about the specific visual effects of these vision disorders. Furthermore, with the increasing popularity of vision correction surgeries such as PRK (photorefractive keratectomy) and LASIK (laser in-situ keratomileusis), our technique could be used to convey to doctors what the vision of a patient is like before and after surgery, using wavefront aberrometry measured pre- and post-operatively.

In addition, by using modeled or simulated wavefront measurements, this approach could provide accurate and revealing medical visualizations of predicted visual acuity and of simulated vision; such simulations could be shown to potential candidates for such surgery to enable them to make more educated decisions regarding undergoing the procedure.

Vision-realistic rendering also has applications in image synthesis and computer animation. It is important to note that the creation of such images based on camera optics follows as a special case of our algorithm. Thus, our approach could be used as a post-process to simulate camera model effects such as depth of field in the generation of synthetic images and computer animation.